A hidden Markov Model (hmm) based speaker identification system using mobile phone database of North Atlantic Treaty Organization (nato) words

---

A Hidden Markov Model (HMM) based speaker identification system using mobile
phone database of North Atlantic Treaty Organization (NATO) words
Shyam S. Agrawal
, 
Shweta Bansal
, 
Dipti Pandey
, et al.
Citation: 
Proc. Mtgs. Acoust.
19, 060019 (2013); doi: 10.1121/1.4800721
View online: 
https://doi.org/10.1121/1.4800721
View Table of Contents: 
https://asa.scitation.org/toc/pma/19/1
Published by the 
Acoustical Society of America
ARTICLES YOU MAY BE INTERESTED IN
On the limits of automatic speaker verification: Explaining degraded recognizer scores through acoustic
changes resulting from voice disguise
The Journal of the Acoustical Society of America 146, 693 (2019); 
https://doi.org/10.1121/1.5119240
Hidden Markov and Gaussian mixture models for automatic call classification
The Journal of the Acoustical Society of America 125, EL221 (2009); 
https://doi.org/10.1121/1.3124659
A Hidden Markov Model based speaker identification system using mobile phone database of North Atlantic
Treaty Organization words
The Journal of the Acoustical Society of America 133, 3247 (2013); 
https://doi.org/10.1121/1.4805213

---

Proceedings of Meetings on Acoustics
Volume 19, 2013
http://acousticalsociety.org/ 
ICA 2013 Montreal
Montreal, Canada
2 - 7 June 2013
Speech Communication
Session 1aSCb: Digital Speech Processing (Poster Session)
1aSCb14. A Hidden Markov Model (HMM) based speaker identification system using
mobile phone database of North Atlantic Treaty Organization (NATO) words
Shyam S. Agrawal*, Shweta Bansal, Dipti Pandey and Himanshu Tayal​
​
*Corresponding author's address: Electronics & Communication, KIIT College of Engineering, Maruti Kunj, Near Bhondsi,
Gurgaon, 122102, Haryana, India, [e-mail]
This paper describes results of an experiment to conduct Text Independent Speaker Identification of large number of speakers (about 100)
using a standard vocabulary of about 23 NATO words- such as Alfa, Bravo, etc. These words in isolation were spoken in a sound treated room
by Hindi natives having very good education in English ( both male and female) and recorded by a three channel data recording system-the
cardioid microphone, electret condenser microphone and a NOKIA mobile telephone. The pre-processed digitized database of isolated words
was further processed to determine 39 MFCC's and their derivatives and used to build an HMM model for each speaker based on all the words.
The HMM model was trained using an HTK tool kit to generate the model parameters and tested using Viterbi algorithm. The identification of
speakers was done in a closed set manner, based on comparison of each NATO word in the model. In addition to correct identification, false
acceptance and false rejection scores were also found. The results show varying performance due to variations in channels, male/female
speakers. The overall identification scores vary between 60% to 70% .The paper gives detailed analysis of results.
Published by the Acoustical Society of America through the American Institute of Physics